Parallels Server Bare Metal + Centos Host - weird log messaged

[futureweb]

Hello,

using Parallels Bare Metal Server 5.0.0 (1102) (2.6.32-042stab39.11) with Centos 6.1 (2.6.32-131.21.1.el6.x86_64).
Within the Host, a few times a day, weird Leg entries appear in the message Log:

Dec 21 05:16:28 icinga kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
Dec 21 05:16:28 icinga kernel: ata3.00: failed command: FLUSH CACHE
Dec 21 05:16:28 icinga kernel: ata3.00: cmd e7/00:00:00:00:00/00:00:00:00:00/a0 tag 0
Dec 21 05:16:28 icinga kernel: res 40/00:01:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
Dec 21 05:16:28 icinga kernel: ata3.00: status: { DRDY }
Dec 21 05:16:28 icinga kernel: ata3: hard resetting link
Dec 21 05:16:28 icinga kernel: ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
Dec 21 05:16:29 icinga kernel: ata3.00: configured for UDMA/100
Dec 21 05:16:29 icinga kernel: ata3.00: device reported invalid CHS sector 0
Dec 21 05:16:29 icinga kernel: ata3: EH complete
Dec 21 05:52:16 icinga ntpd[1203]: synchronized to 193.67.79.202, stratum 1
Dec 21 05:52:33 icinga ntpd[1203]: synchronized to 192.53.103.104, stratum 1
Dec 21 05:58:43 icinga kernel: ata3.00: exception Emask 0x0 SAct 0x2 SErr 0x0 action 0x6 frozen
Dec 21 05:58:43 icinga kernel: ata3.00: failed command: WRITE FPDMA QUEUED
Dec 21 05:58:43 icinga kernel: ata3.00: cmd 61/08:08:c0:c8:54/00:00:00:00:00/40 tag 1 ncq 4096 out
Dec 21 05:58:43 icinga kernel: res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
Dec 21 05:58:43 icinga kernel: ata3.00: status: { DRDY }
Dec 21 05:58:43 icinga kernel: ata3: hard resetting link
Dec 21 05:58:44 icinga kernel: ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
Dec 21 05:58:44 icinga kernel: ata3.00: configured for UDMA/100
Dec 21 05:58:44 icinga kernel: ata3.00: device reported invalid CHS sector 0
Dec 21 05:58:44 icinga kernel: ata3: EH complete

Everything is installed on our SAN - no local disks. No other Server complains about any "HDD problems".

On the Bare Metal the message logs doesn't say too much - only this:

Dec 21 05:14:21 vix10 kernel: [681804.492022] /drv/ lin_async_device.c:214 host_wait_work_complete_async_device (08:00) finished with timeout.
Dec 21 05:14:22 vix10 kernel: [681805.492204] /drv/ lin_async_device.c:214 host_wait_work_complete_async_device (08:00) finished with timeout.
Dec 21 05:14:23 vix10 kernel: [681806.492199] /drv/ lin_async_device.c:214 host_wait_work_complete_async_device (08:00) finished with timeout.
Dec 21 05:14:23 vix10 kernel: [681806.639032] qla2xxx 0000:10:00.0: scsi(0:3:151): Abort command issued -- 1 2002.
Dec 21 05:56:37 vix10 kernel: [684340.329041] /drv/ lin_async_device.c:214 host_wait_work_complete_async_device (08:00) finished with timeout.
Dec 21 05:56:37 vix10 kernel: [684340.696030] qla2xxx 0000:10:00.0: scsi(0:3:151): Abort command issued -- 1 2002.

any ideas what's the problem here?

Thank you, bye from snowy Austria
Andreas

[MichaelB1]

Did you ever come up with an answer? I'm having this issue now.

[futureweb]

unfortunatley not, but we dropped production use of bare metal for the moment and are sticking on virtuozzo containers, as we were seeing weird crashes by using centreon monitoring (kernel panics that seems to be disk issues). As Centreon does quite some heavy disk writes, we assume it might be a reason as we saw huge amounts of IOP's on the root system too only from this one VM.

We migrated this VM to a CT, since then, all works fine (using the same bare metal install though).

Next time we might open up a support case, as we haven't done that so far. But as we are moving from Fibre Channel Storage to ISCSI Storage, we wanted to wait on completion of that migration, before trying it all again.